Questions about a game in progress
 

Hi to all!

I have started playing Dom 3 a few days ago (after playing a bit of Dom2 way back in 2004, but not much since my PhD took most of my time! But now it is almost over!) and I am looking for some advice since I can be considered as a noob (I have to add that my usual playing style in strategy games a la Alpha centaury for instance, is to turtle and be the best in research. This is usually what I try to emulate in Dom3, but I am not sure this is an optimal strategy).
Indeed, my first three games ended in disaster after a few turns while I was experimenting with some new nations and imprisoned pretenders on the Aran map; Then I decided to go back to the basics in order to address two of my main problems, namely, province defense and magic site searching. To this end, I build on Ulm middle age, with a rainbow master alchemist (my mistake was to take Nature 4 and only F1), Order 2, Productivity 2, Drain 3, Misfortune 1, and a dominion of 5. The idea was that; since Ulm units cost a lot of resources but are very resilient, a few would be necessary to expand and gold would not be an issue with Order2 so that later on I could buy a lot of province defense. Also, the rainbow alchemist was supposed to be my site searcher early on.
So, I started a game in the Aran map against 5 Normal Random AI, with Indy strength set at 5 and magic sites at 50.

Everything went according to plan initially and I was able both to expand and to find magic sites while keeping a lot of gold in my treasury. Soon, I was able to build a second castle and I found my neighbor, Pangea then my second neighbor, Agartha. Now, I am in winter of my third year, I have just beaten Pangea and I am on the verge of building my fifth castle. My treasury is 9137, with an income of 3019 and an upkeep of 863. I think I am at a turning point since I am going to clash with Agartha which is the second most powerful nation (as many provinces and forts as me, larger army, but they are way behind in research.). Although my armies were sufficient to now, am I afraid that my lack of magic might be detrimental in the long run so I wonder what I should do. I have started to build a few items to boost my pretender (like ring of sorcery, crystal coin, earth boots...) such that he is now F3 A2 W3 E3 S6 D3 N5. When I have a sufficient amount of astral gems, I will build a ring of wizardry. Then, what should I do? The problem with Ulm is that my master smiths are only F1 E2 so that my magic is very limited. I have discovered a province where I can buy Jade Sorceress which will help me in getting Nature magic. But what else? As to my gem income, I am not sure it is large enough: F+6, A+2, W+1,E+8,S+1,D+3, N+7.
I have a mage cast Gnome Lore every turn, and my only mage with F2 casts Augury. Is it enough? Basically, I can't use my pretender any more since he is forging items. But I don't have access a lot of astral pearls, but I can't forge Clam Pearls since I don't have a lot of water gems either.

By the way, how do you usually search for magic sites? Here, although I had a pretender dedicated to this, I don't feel like my gem income is sufficient...So, I wonder how people search without a rainbow mage. Do you dedicate a large part of your mages to remote searching?

As to the game at hand, any ideas are welcomed. I don't have a lot of experience with high level magic or items, so I am a bit at loss of what I should do next. Thanks for your help!
I know this kind of question depends on the specifics of the situation, but any suggestion would still help me.

Re: Questions about a game in progress
 

I think I might go for more Death magic since I have 71 Death gems. As soon as I have finished Conjuration 6 (2 turns) I could summon spectres which will give me access to more magic.

Re: Questions about a game in progress
 

Remote spells make spellcasting much easier, so they are often used. Manually moving a site-searching wizard around can be worth ti too, though, especially if he's a rainbow pretender or a mage with many paths. You just have to make sure your enemies can't attack him.


Ulm Master Smiths can cast some useful spells. In Evocation: Magma Bolts, Blade Wind, Magma Eruption; the ones with Astral random can also cast Gifts from the Sky, those with Air can cast Rain of Stones. Conjuration gives few important battle spells: Phoenix Power and Earth Power let your E2 and the rare F2 mages cast better spells, and is especially good for all those E2 mages who can now cast Blade Wind without carrying gems to battle or wearing Earth Boots. Earth Boots AND Earth Power let them cast their spells many more times before fatigue makes them collapse, but the reinvigoration from Earth Power helps in there as well. You might also consider Destruction, an alteration spell that destroys armor, if you can ensure that the Smith casting that has good enough precision to hit your enemies and not your own troops.

You might be able to summon some mages. Conjuration has some Nature (Lamia Queens, Ivy Kings) and Death (Spectres, Mound Fiend) summons, and Water can get you Sea Kings and Naiads in Conjuration or Unfrozen Mages in Enchantment.

If you have any Master Smiths that have S1, be sure to set one to cast Arcane Probing, to look for further Astral sites.

You do have a laboratory in all of you castles, right? Even Ulmish forces need support mages, and Master Smiths can do that very well. In addition to bashing your enemies, they can cast Legions of Steel and Strength of Giants, especially the first one is very good for Ulm.

Re: Questions about a game in progress
 

You are working on a ring of wizardry, and you probably don't have a skull cap yet, but sooner or later you should summon a Mound Fiend (D3) - he can then be given the job of summoning spectres to give you more varied magic (D?2, where ? is WESD). S will allow you to sitesearch all astral via spell, W to map all water (on land) with the bracelet.

Early in the game (which you are not), it would have been better to use the cheaper Revenant (and give him a Skull Staff so he can multiply http://forum.shrapnelgames.com/images/smilies/laugh.gif) to cast dark knowledge.

You say you have one mage sitesearching fire. If you have lots of provinces left that need to be searched, your pretender might want to forge a flaming skull (FD req) that you can give to boost another fire mage to cast augury.

Your paths allow you to contact Naiad (well, one more water, but that is easily forgeable) if you have a surplus of water gems, giving you a unit that can hammer out clams, sitesearch water/nature, or contact other Naiads.

How I usually search for sites (in long games):
* If I have a rainbow pretender, I use him to search for sites the first 10-20 turns after he is activated (awake or dormant, never imprisoned for this), then he is usually retired to forging/rituals.
* If I have any decent national mages with multiple paths or high single paths, I have them out searching manually. Priests too once I have a surplus of them - you'd be amazed how many Ancient Temples with 1 astral gem income slips under the radar if you just use mages.
* Sitesearching spells of all types have high priority in research. Especially if I have mages with D1 or S1, getting that death and astral gem income going early is important. A good starting astral income will help you when sitesearching other paths where you have a low income as you can convert the astrals as needed - it does pay off. The earlier you get things sitesearched, the better.
* Construction 4 to create cheap and useful pathboosters, and to start forging skull mentors with the death gems found, has a fairly high priority as well.
* Use death and nature magic via conjuration to branch into areas of magic your nation doesn't touch at all. If your nation doesn't have good death or nature mages, your pretender should probably have these paths unless you are going for a blessing strategy or a dedicated SC. Sure, you might find good independent mages - but then again, you might not, and having a guaranteed way to get mages in the mid-game to bootstrap yourself into a power in all (or most) of the magic types is important.

Ulm is probably the least magical nation around, but even so, assuming you plan for [get some death/astral income going early on] it you should be able to master at least FWESDN in longer games. (A is harder and B is not going to happen except in Black Forest) Then again, you might get uncommonly unlucky in sites, but, on average, you won't be.

Re: Questions about a game in progress
 

Ulm can get 1 in Astral and Air, and 2 in Fire, nowadays, with just Master Smiths' randoms.

Hidden in Snow, Enchantment 6 and W3D1, is expensive at 55 gems but gives you up to two Unfrozen mages who the manual lists as having four randoms. That would be very nice if you had more Water gem income.

Re: Questions about a game in progress
 

Blade Wind at Evo4 and E3 is your friend.

I usually fit every support mage with Earth Boots, bringing them to E3. Just a couple of mages casting Blade Wind can seriously damage a lightly armoured army (I suspect Agartha fits that description), especially one that relies on superior numbers.

At least that's from my limited experience.

Re: Questions about a game in progress
 

Or you can just cast earth power (as mentioned) bringing them up to e3 without wasting gems.
If i was you i would (yes i do suggest this to everyone http://forum.shrapnelgames.com/image...ies/tongue.gif) i would use my pretender to summon a death mage and then use him to crank out black servants (cheap, stealthy, ethereal, lifedrain) for raiding and spectres (basically the same but more expensive and with magic for buffs) to hunt small armies (or if you send groups they could possibly do better) and equip them with cheap items (protection, strength boosters and vrs most troops fear are all top priorities) from smiths and some different items (special shields, horror helms, luck pendants ect) from summoned mages or your pretender.

Re: Questions about a game in progress
 

Thanks guys for the input!

I see that my early siteèsearching left much to be desired.
I have one master smith with S1 so I have ordered him to auto-cast arcane probing. Better late than never!
Also, I am going to summon some spectres and mound fiends very soon I guess. Maybe I can start summoning black servants also...

Re: Questions about a game in progress
 

Sure cast Earth Power. But we're talking Ulm here. Give every combat mage Earth Boots, it's only 5 gems apiece. With Earth 4 and reinvig, you can cast more Blade Winds before fatiguing out.

Re: Questions about a game in progress
 

Only summon 1 death mage with your pretender though, he has better things to do. Black servants make very skilled raiders, heres an example of (imo) a nice one:

no weapon (built in lifedrain ftw)

charcoal shield (your main damage dealer, works wonders vrs non jotun pd/troops)

horror helm (combined with fireshield its a very nice way to rout the enemy sooner, especially if you used in groups)

black steel full plate (great prot to take hits, need i say more?)

boots of giant strength (boots lifedrains damage, very helpful)

luck pendant (doubles your survivability, almost essential on any combatant without astral magic)

amulet of antimagic (helps you resist those deadly banishings, if using more than one black servant spread them out so they dont all get hit)

and there you go, for 5 astral, 13 earth, 3 fire and 8 death (thats off the top of my head and if your astral/earth/shield items are made by smiths with hammers) you have made a raider who takes almost no damage from attacks (not all attacks hit, mundane hits only have a 1 in 8 chance of getting past luck and ethereality and even then they have to get past your nice protection), is resistant to magic, can heal himself, has fear AND damages anyone who attacks him.

Re: Questions about a game in progress
 

Yeah, what is up with that spell?

I cast it three times and got:

2 mages, 5 improved unfrozen, 20 normal unfrozen
2 mages, 7 improved unfrozen, 18 normal unfrozen
0 mages, 3 improved unfrozen, 6 normal unfrozen.

The last casting was during Winter, and boy did that suck.

Anyways, the Mages are random, the worst one was 2 Earth and 1 Death, but the other 3 had 6 or 7 picks each, distributed between Earth, Water, and Death (looks like up to 3 in each path, I got a 3 Earth guy, a 3 Water guy and a 3 Death guy with 1 or 2 picks in the others).

Individually, Unfrozen have better stats than Wights, but they tend to fight with stone spears and leather armor, which makes them somewhat inferior. The Unfrozen Leader however, is pretty exciting. Marginally better stats than a Bane and his disadvantage of coming with worse armor is just 5 Earth gems away from being solved.

-Frank

Re: Questions about a game in progress
 

Ive only cast it once and I got 1 unfrozen leader, 2 very nice mages, 1 poor mage, 2 improved unfrozen and a bunch of regulars.

Re: Questions about a game in progress
 

In general "turtle and research" is a tactic which would work best on
A) large maps
B) few players
C) high setting on independents

Any of those can work, or all of them. Any opposite settings (small maps, many players, default or lower indepts) would create a rush game which would be hard for you to survive. Dont despair though. If you are good at turtle and research then you could be a welcome surprise in the large games.

Re: Questions about a game in progress
 

Perhaps the spell has worse results during the winter and best results during the summer... it might also be complete random results. If I wasn't at work I'd do some quick tests.

Re: Questions about a game in progress
 

Ah, that reminds me of something else. If you forge a Dwarven Hammer (E3), anyone having it while forging will have a forge bonus 25, the same as your Master Smiths, and it stacks with your Smiths' ability. Can you say no to 50% off of all items?


Black Servants aren't mages. They are ethereal, however, and as most undead they have encumberance 0. That means that they are hard to kill, and won't suffer from high-encumberance armors (their encumberance will STAY at 0, no matter what armor they wear). Black Servants equipped with Black Steel Full Plates and Black Steel Helmets are VERY hard to kill. Another very useful item is Pendant of Luck (S1), having 50% chance of negating anything that might further damage them, and of course Amulet of Antimagic (S1) to protect them against enemy priests. They could also benefit greatly from Lead Shield (E1 or E2), which also increases magic resistance, especially as they can ignore its encumberance.

Re: Questions about a game in progress
 

Who said they were?

Re: Questions about a game in progress
 

Ok, my game is going well for the moment, even though I guess it is going to be harder now that I am at war with Tien Chi, Caelum and Agartha at the same time.

I had forgotten how much micromanagement was involved in the middle to late game: it is a pity mages can't autoforge the same item every turn...And assigning the ten lightless lanterns you forge each turn to 10 mages is not funny either. But I guess that's the downside of having so many possibilities.

one thing however: I was casting something like 2 Haruspex, 3 augury, 3 gnome Lore, 3 arcane Probing and 2 Dark knowledge a turn, and I found that my success rate was low. Then I took a closer look, and it turned out that the mages were all searching the same province on a given turn, even those that were casting the same ritual! I guess this is a bug, or I am just unlucky. Hence, I have stopped casting each spell more than once per turn. But it's a pity.

Re: Questions about a game in progress
 

It's a bug. It seems none of us remembered to mention the bug... http://forum.shrapnelgames.com/image...es/redface.gif

Re: Questions about a game in progress
 

ok. So, basically, it's pointless to autocast the same spell multiple times.
It is kind of annoying...

Re: Questions about a game in progress
 

Yeah, it is. I mean, it might be bad strategy to autosearch as much as you can, but I'd like to be able to decide on my own bad strategy (esp. in SP) without the game making the micromanagement help only available if I limit my searching...
(and as a result, I end up only casting each spell once per month, because the monthly casting is so much more convenient)

Re: Questions about a game in progress
 

You could still have one type of search going on monthly. But when you have enough free mages and gems, those can be used to manually search other provinces. The monthly spell still tells you its target so you can adjust your other searches.

It's kind of like an indexing service on a computer. Usually run in the background, slow but steady with little use of resources, but when needed, just activate it and put some more CPU time to it. It just depends. If you want results quickly, you have to micromanage. If just want results, let the mages do it themselves.

Re: Questions about a game in progress
 

It's very annoying. But it's nice to be able to autocast it at all. I usually have one autocasting each search spell and do the rest by hand, making sure to leave the next province for the auto searcher.

The auto search spells pick from the top down, so I just start the manual searching from the bottom up.

There are some other weirdnesses with autosearch, but I haven't pinned them down yet. It will ignore provinces occasionally, for reasons that I don't yet see.

Re: Questions about a game in progress
 

what do u mean by "the auto spell pick from the top down"?
alphabetically?

Re: Questions about a game in progress
 

By province number -- it'll go down the list from top to bottom as you see it when you hit F1. The only problem is that it may target a newly captured, and therefore vulnerable, province before it targets "safe" provinces

The other weird thing with autosearch is that it will target your own capital. I haven't seen it target a conquered capital yet.

Re: Questions about a game in progress
 

Yeah, but if the goal is to help the player reduce micromanagement, then it should really help - like, let the player choose how much searching he wants to do, and start helping from there.

Right now, the site-searching part of the monthly casting is saying, "OK, I'll help you by auto-searching for sites - but only if you want to search one province per path each month. Any more than that, you're on your own", and there's no good reason for this, other than "that's how the computer does it" (which is exactly what I tell my students not to do: the program designer decides what the computer will do, and implements it, rather than implementing something and describing the feature from the result).

Re: Questions about a game in progress
 

It will only sometimes target your own capital. It may be by nation, but I'm not sure. I know some games it has and some it hasn't.

It will target the next province in the list (by province number) that hasn't been searched when the mage finishes casting his search spell. If you have another mage casting the same search spell on that province, but he casts after the autosearcher, that province hadn't been searched when the autosearcher chose his next target, so he will cast on it again. That's why I do manual searching from the bottom up.

Worse, anything that happens after spellcasting isn't considered. Newly conquered provinces aren't possible targets, but if you lose a province already targeted it will still be searched.

This isn't going to be an easy fix. To do it right, they'll have to make some significant changes in the order things are done in turn processing.

Re: Questions about a game in progress
 

While we're on the subject of the "Nation overview" (F1) screen - am I the only one who finds it annoying, in a large game, to hunt for a province that's just described by it name in the messages screen (typically, "RandomMage cast some searching spell on province BlahBlah and found 2 magic sites", only the "go there" button takes me to the stupid mage rather than the juicy province and now I have to look through the pages upon pages of province to find it)? Some option to search this screen would be a big help - even the spell lists have filters by path now, this is at least some help.

I have.

Re: Questions about a game in progress
 

It SHOULD be possible to search a province you lost, and impossible to search provinces you won't conquer this turn - that happens when you search by hand and an enemy does a surprise attack.

Re: Questions about a game in progress
 

But that's not what happens.

The auto searcher can target a province you've lost. That spell will be cast next turn. You couldn't select it to cast a search spell into if you were doing them manually.

Re: Questions about a game in progress
 

I see. It makes sense, though. If the mage is reset to cast the same spell immediately after the previous spell has ended, before provinces are conquered/lost, and even before magic gems have been received... well, that explains the other quirks in the monthly casting.

Re: Questions about a game in progress
 

Yeah, it makes sense from a programming point of view.

But it shouldn't work that way.

Re: Questions about a game in progress
 

While I'm not happy with auto-site-searching only being good for one search per path per turn, it _does_ sometimes have an excuse for targetting capitols. With a number of hand-crafted maps, your capitol could be right on top of a hidden
magic site. This lead to reading the .map file to see if there's something good in your capitol, and then manually searching for it. http://forum.shrapnelgames.com/images/smilies/wink.gif

Re: Questions about a game in progress
 

I doubt it's anything that calculated. My impression is that it sometimes searches capitals and sometimes skips them. It also occasionally ignores other provinces.

My suspicion is that the capital problem is by nation. Some nations capitals are always searched, some are never, but I haven't confirmed that. Which would mean a flaw in whatever it's using to determine what is a capital?

Re: Questions about a game in progress
 

I've noticed that auto-search seems to skip any province that's been searched at -any- level in that path, even if the previous search level is less than 4 (let alone 9). It's a bit galling if, say, you're EA TC.

Re: Questions about a game in progress
 

This one doesn't seem like a problem to me. It just means that the spellcaster acted before the battle happened. (The idea being something like that rituals have to happen exactly the same way everytime, and can't be stopped once they're started, which is often the way powerful magics are treated in fantasy.) Sounds less like a bug and more like a calculated risk.

Now the fact that the autocaster will target vulnerable provinces is a problem because it doesn't model minimally prudent strategy, which would be to target provinces you aren't likely to lose before one's that you are, unless you have good reason to expect a strategically significant gain from it that's worth the risk.

Re: Questions about a game in progress
 

Except that nothing else works that way. Normally you can stop his order, change it to something else, change it back etc.
But in this case, you can still change the order, but you can't change it back. It doesn't fit. It's obviously an artifact of the order the program executes orders in. It's a bug.

You can come up with justifications for it, but they're just after the fact justifications.

Well, if you can come up with a good, simple algorithm for "provinces you aren't likely to lose", combine it with the other obvious strategy of targeting provinces more likely to have sites of that path, weighted to keep everyone happy, I'd love to see it.

Generally, by mid-game I've given up trying to prioritize site searching and am just going up or down the list, because it's too much work for me to figure out which provinces are best to search.

I do find it amusing that you're not bothered by it targeting provinces you have lost, but are bothered by it targeting vulnerable provinces.

Re: Questions about a game in progress
 

Are you saying that in a turn *after* the one in which you lose it it targets it? And that you can't change the order manually? That's goofy.

I thought you were just saying that it targets a province and searches and then a battle is fought and you lose, all on the same turn...

You're also right that to fix the thing I'm talking about would be nearly impossible. I guess the closest you could get would be to have it target province numbers closest to the capital's #, and hope that #'s correspond to proximity a little...

Re: Questions about a game in progress
 

You can change it manually, but once you do you can't set it back to the lost province.

You could, in theory, have it first search provinces that only neighbor provinces that you own, but I'm not sure it'd be worth the effort.

I settle for getting the current version working.

Re: Questions about a game in progress
 

Ok, I understand the glitch, but why would you ever want to set it back? In hopes that you'll recapture and get a site search in a single round?

Re: Questions about a game in progress
 

I'm sure the glitch is not permanent... waiting one turn is not that big of a deal.

Re: Questions about a game in progress
 

Hi all, first post.

This is a pretty interesting problem. I'd suggest a simple fuzzy logic algorithm.

The basic idea would be to calculate a score for each province you control, based on n criteria. Each criterion would award points for that province, depending on how well the province matches it on a scale of 0 to 1. Each criterion would also have an intrinsic importance value, which would be used to scale it against the others. The actual points received would be (match) * (importance).

When the scores are calculated, the provinces are sorted according to score, and the highest scoring is selected. Provinces searched to 4 or more in the relevant paths or targeted by an overlapping search are excluded.

In my programming experience, these kind of things work best when you concentrate the weighing on a few things, and keep the rest as "tiebreakers". I'd suggest the following:

1: Ratio of friendly neighboring provinces versus enemy controlled ones. This would cause a tendency to avoid frontier provinces, or provinces you have captured with raids.

2: Previous searches. This would pick unsearched provinces over ones with 1, 2 or 3 level searches. A bit trickier to calculate for Aschaic Record, but possible.

Tiebreakers:

3: Probability of magic sites. Wastelands would get targeted over farmlands.

4: Probability of magic sites for this spell. Gnome Lore would pick mountains over other terrain and so forth.

5: Number of other sites already found. IIRC there is a maximum number of sites that can be found in one province.

It's not a perfect system, and never will be, but it'd be a lot better than going through the province list from top to bottom. Assistive features such as the auto search are not meant to make real decisions for the player anyway, just to remove the chore of making the trivial ones.